Chemistry. -- "T/w Condition of Motion of thIJ Molecules 'Ï'tl

Space." By Pl"Of .. 1. HÖLf,SEK[N. (In collaboralion with Messl's .

CHIt. VAN LOON, DERX , and HERMANS).

(Communicated at the meeting ' of October 29, 1921).

In apapel' 011 tbe configuralion of Ihe tarlarie acids (BÖESEKEN and Cao ps) 1) Ille concIllsion was drawn from ' Ille bebaviour of these acids and their esters towards bork acid, that the carboxyl groups, both when subslituled and when not suoslituled, were as , fal' 'apart as ' pORsible.

The amide!! alld ethyl-amides, however, presented deviations, .so that this simple assumpt,ioll did not sufliee. The great influerice which these subslanees have 011 Ihe conduclivily of boric acid, poillts 10 a fa:vourablè sjtualion of the hydroxyl groups wilh regard io eàch ol he .. , which can only satisfactol'ily be aeeoullïed for,w'hen there al'e attmctive acti·ons between these gl'OUpS and Ihe'Subsliluted or non-substilutedamide groups: '

Flll'ther for a cOmIJal'alively large numbel' of a.glycols ·no inflil­. ence 011 ' theconJuetivity oflhe bo .. ic acid wa's found jthese OH­g'roups must, I 11 el'E!fore , lie rat, apart, which could again beexplained

' by mui,ual re·ptilsion . At any rate il appeal'ed from these investiga­tions thai ' aloms that are not directly bound 10 eacll other, ,afso exel't an action on each Ol 11 el', and we may ' expect that the stit'te

. of ' eqüilibriurriof ft molecule will be the ' l'esuÜ' of all . the fe'rees in that molecule, bOlh the atli'aclive Rnd Ihe repulsi,'e o'nes,

I have all'eady poiuled olll,S) , thal evel'ychallge ,in Ihe posilionof the aloms, however slighl, must give rise 10 re-a .. rangemEmts, thl'ollgh whicli the position of -all 'Ihe parls of the molecule willnot remain exactly Ihe same.

In the molecule-halves of active and i/lacti~:e ' tal'larie acid e ,g, ' the compal'able gl;oups will ' no 'Ionger be perfeelly iJentical j hènee the pl'inciple of the oplical supel'posilion canllot be valid ,

In consequence of the action of these fOl'ces, molecules which up to nowhave been cönsidel'ed ai! symmetl'Ïcal, as the anli-tartal'ic

1) Versl. Kon, Ak, Wet 29, 368 (1920). Has not yet appeared in English. ~ .... .. 29, 562 (1920), .. ..

199

acid, will no longer ' lIave minor images thatcovel' each other, ;whieh ,can ' easily be demonstrated by means of the well-knowlI eal'bon modeis,

Only two positions of the anti tartaric acid are symmetrical, viz. thal in which the equal groups lie on lhe same side of the axis of lbe cent ral C-atoms as close

. as ,possible to each other and th at in whicb they lie on bOlh sides as far erom each other as possible. TI would be qllite accidenlal, that one of these positions should be the most stabie state of equilibrium under all circumstances.

This seems in contIiet witll eÄpel'iellce. We SIlOllld, Ilowever, bear ill mind (he following cOJlsiderntiolls , In view of the limited numbel' of isomel's it lias all'eady 10llg been assnmed thai tlle molecule

,halves can move rOllnd Ihe sillgle bonds as axes in opposite direct.ion Ol' wilh diffel'ent \'elocities.

When wesllpplement tllis necess:\ry condition with the hypolbesis that these movemenl,s also continually take place in tliis sen se, that Ihe most stabie posilion of all Ihe posiliolls possible at Ihe momel'lt :wiil occüt , most freq uenll)", and that, thel'efore, the molecule execules ir'regnlal' , rotat.iolls Ol' Osei Ilat iOlls I'OU nd th is posi tion, there is no longel' MIJ' contradiclioll wilh experience.

When we eOllsider tllese rolaliolls I'olllld the bond ófthe centrAl carbon atoms of tlle allti-Iarlal'ic acid, we see that in one full rotalion of one part wilh regard to tlle olher, the molecule will twice pass a symluett'ic position.

At these momenls tlle molecule is oplically inaclive, and since now the cllance lias become equally great that Ilie siabie asymmetric­position will be 'I'eached by a movement iri Ihe dil'eclion of the hands of a clock Ol' in opposite dil'ection, a contirlllal racemisation will take place. , ',

Thollgh the state of: eqllilibrium is aSJm~etric, we shall pl'obably Jlevel' obsel've I he optical act,i v i Iy i Jl I iq u id Ol' dissol ved state.

This dYllamic rcpI'eseJllatioll, wlliell is fOI'ced "pon liS by Ilie inacli\'ily of ' lhe anli-Iarlaric acid, applies, of COIII'Se, 10 all salul'aled

" molecules, Om' ûbservaliolls of Ihe bella\'iollr of tlle ft glycols 'towai'ds boric acid and acelolle call be explained most simply in Ihis way.

Tlle sat u rated J1on-cyelic cc-glyèOls do 1101 inerease t he cond llcti vity of the borie acid; we lIave drawn the conclusion from Ihis that the lIydroxyl-grollps al'e 1101 favollrably sil\1ated, and have accollnted fOl' this by Ihe assumplioJl that the OH-gro\1ps repel each oth~r,

thro\1gh which Ihey can gel as fal' as possible frol.! each -oUIer owing 10 the mobility of the molecule.

According IQ the dynamie l'epl'esentatioll this position will be the

200

most stabie state of tlle molel'llle, but all the othel' posilion8 wjll 11.180 be realized, albeit less fl'eqnently, and the mOl'e rarely as the state deviateR more eortSid~rably from the most slab Ie,

Of the u-glycols there are eyclic aftet.one-compounds known, whieh are formed with excess of acetoné, to which ti. Iittle HOI Ól' H,SO. has beeu ádde'd, and whieh a/'e stabIe, when ,the~'e is not á trace of these laRt catalysts present.

TIlere is every reason 10 assnme th at Ihere is the greatest chance that ' these compounds are formed when the OH-groups al'e sitnated in the plane with the O-aloms 10 wh iel. they are bound, and on the same side bf them, When the OH-gl'oups repel each othe/', t.his will seldom take plaee, but by taking a \'ery gl'eat excess of aeetone, we shal1 yet be able to obtaÎn a pretty lal'ge quantity of the cyclic compound, whieh can be retained by /'emoval of tha ('atalyst, and whiclt ean be freed from the aeetone, by fractional distillation ,

When the OH-groups cannot get into tlle most fa"oUl'able position, or near it, as in some eyclic trans 1 ,2-diols, where the free rotation th/'ol/gh the ring is prevented, no acetone-eompounds will be formed, This rlOw ha!" been found, and this renders a firm support tó t.his dynamic representatioJl (see below),

The satn\'ated glycols can, however, always aBSnrDe the most favourable position, thongh it oe only ral'ely; the relative frequency of this case may be judged from the state of the eqllilibriillll:

glycol + acetone ~ acetone compound + water.

By determination of tlle constant of equi1ïbriunl:

K. _ Cglvc.ol X CA.CP.t.one

t' - Ccomp, X CH,O

a measure is sllpplied fOl' this fl'equenry, and so , ,also á mea~ure fOl' the situation of the hydl'Oxy I gl'oups in the most stabie position of thé molecule,

Hel:e' fol1ows a summal'y of the constants of equilibrium of s~me glycols: 1)

1) In his thesis for the doctorate (Delft 1919, p. 59) "over de Stere.ochemie der' cyclopentaandiolen 1.2 en hydrindeendiolen 1.2". Mr. CHR, VAN LOON was th{ first ' to point out that the study of thése equilibria would probably' bring to light minuter differences in the configuration of the molecules than the borie acid method. The measurements were carried out by Mr. p, HERMANS,

201

K180

ethylene glycol 0. 14

propane diol 1.2 0.44

• 1.3 0.026

:x-monochlorhydrine 0.28

glycerine 0.77

cis-cyc1ohexanediol 1.2 0.15

trans-"

1.2 0 .000

The lltudy of the eyclic glycols has supplied an important support 10 Ihe efticlency of Ihe dynamic representation.

In the cyclo-pentane-diolg the five carbon ato.ms lie in one planè almost wilhout tension wben we assume thaI tbe dirertioJls of affinit.y make an angle of 1090 28' with each other. In tbe ci.s-isomers tbe OH-groups lie almost in one plane with t he C-,atoms, 10 which they are bOUlld, he'llce veJ'y favourably for bodc acid and acetone, as has appear.ed from V A~ LOON'S investigafion_

In the tJ'ans-diols the position is ver] unfavouJ'at>le. In eonsequence of the five-ring the position of these OH-grollps can never become favollJ'able without the ring being broken, Indeed lthere was DO.

qllestion of Ihe formation of an aeetone compound , The absence of tbe aeelone-eompound in Ibis tJ'ans-diol and in

the tl'ans-bydrindene diol 1.2 is also the proof th at when lhe OH­groups al'e prevented fl'om getting erom time to lime info. 0.1'

almost inlo o.ne plane, no ft ve-ring is fOl'med_ (see ,above),

At thé time it was fonnd that cis-cyelohexane .diol 1 ,2 has no intluence on Ihe condllctivity of Ibe bol'Ïc ac.id, This bad led me 10

eonrlude all'eady then that Ihis sil.-J'ing must be aSBumed to ha.v.e a certain suppleness, Ihrollgh which the hr dl'oxyl gJ'oups wo.uldg.et somewhat fllJ,thel' from eaeh othel' Ihan would have been ex,pected, if the six atoms ,of Ihe ring lay permanently in aplane,

The equil ibrium: glycol + ace.tone ~ acetone compound + water

COJ'l'obol'aled on one side thiB view, as .its positiaD ,is .almast as unfavourable as wilh the ethylene glycol (see above),

011 the oUler hand it cOl'l'obol'ated our dynamie view: The acetone eompollnd is fiormed; hence the hydl'oxyl .gl'on,ps must ,nowand then lie in Ol' almost in t.he same ,plane as ·tbe adjacent C-atoms.

202

A complete I'otation, as is tlre case witb the saturated non-cyclic glycols, is not possible hel'e, as this would mean a cOJltinual transi­tion from cis- int.o tmns-diol, and viee-versa, whi c lr does not take place ; besides, t.he tra ns-diol g i\'es 11 0 tmce of an acetone compound,

How can we now form an idea of the movement in tlre cyclo­Irexane diol?

In 1890 an extensive s tudy was published by SACHSE 1) on the

position of the C-a tom s in tlr e saturat ed ring systerns starting f~om

the premise tlr a t tlr e direct. ions of a tlinity ' form angles of 109°28' with each otber, He pl'oves fOl ' tlre tlrree-, fom-, and fjve-rings that lhe atoms must be s itllated ill o li e (lIane, and that a tension is inevi tabIe, w hic lr is, 1Iowe\'er, i nsign i fi call t for Ule ti ve-ring, (VON HAYlm'S ring ten sion ), He shows fo l' tlle s ix-ring th at the molecule ean escape a lellsioll in two ways, Fil'st of all the C-atoms can lie three in one plane, and three in another, so that lire bonds lIetween the C-aloms form a zig-zag line , lt is, ho wever, also possible Ihat {'our atoms lie in one plane, and two (e,g, 1 and 4, 2 and 5 Ol' 3

and '6) ontside it. :. Tobring the molecule out of tlre firs t position, a certain resistance

must be' su rmounted; tlre second position is pliable, It has been pointed out on different s ides that these positions of

Ille molecule conld not form t.he image of lire stabie situation of the atoms, becallse the numbel" of iso mers, and particularly of optical isomers would tlren have to be much g reater ,

ASCIIAN ' ) , ho wever, rema rk s, th a t these positions w'e possible when they al'e considel'ed as the succestl ive phases of a system in motion ,

This view, indeed, has been cOHIil'lll ed by . ollr boric acid measure­ments and the s tIld y · of the acetone-cornpoullds.

Wlr en we accept Ihe view thai tlre molecules of tbe six-rings move tlrl'O'ngl'l space as undulatory surfaces, the OH-gl'oups of Ule cis-diol get fl'om time to time into tbe same plane witlr and on the

' same side of the C-atoms to which Ilrey al'e attached, and Ihe freqllency of thi s OCCUlTence will depelld on lire most s tabie state of the molecule.

It is clear that in Ihese undlllatiOlls repeatedly a symmetrie position is passed, so that, though tbe most stabie situation is an a sy mmetric one~ the se paration of optical isomers in liquid st.ate is not vel'y weil . possible; fOl' racem isation wilicontillllally set in fOl'

1) Berichte 23, 1363 (1890). Zeitschr. f. physik. Chemie 10, 228 (1890). 2) "Chemie der Alicyclischen Verbindungen", pag. 328-331.

203

the same "eason as with the movements of the allli-tal'taric acid, In the tmnii-cyclohexano diol the OH-gl'oups always remain vel'y far apart; indeed, we havo 1I0t. been able 10 oblaill all acetone-eompound, nol' lias an illflllence beon obse l'ved 011 tlte cOllducti,' ily of Ihe bol'Ïc acid,

Anotltol' contil'matioJl was furni shed by the cis- and tmnscyclo­heptane diols,

WheJl SACHSt.;'S ca lculations are applied to them, the OH-grollps, both in the eis- alld in the lrans-diol , get ill Iheil' movementsiJlto, Ol' neul'ly inlo the same plane \Vilh t1- lId on the sa llle side of ·the C-atoms, which they adjoill, Now acet.one (~ompollnds have actllally been oblaiIH~d, aJld increase of the condllctivily of bOl'ic acid has been observed, both in the cis- and in the tl'Ons-diol; accordiJlgly our measllremenl s I1l'e in perfeet harmoll y with what is to be expecled in Ihe movernenIs of Ihi s sys lem as nndulatory sUl'face,

According lollli s view tltere is, tll ere fore, no lension in Ihe six­and sevenl'ing-syslems, even less Ihan ill Ihe fi ve-ring, and in facl, we knew this already from Ihe measuremelll s of Ihe heals of com­bllstion of I Ite cyclu-paraffi /l S. Tlte ve l'y accl11'ale delel'miJlations by S'I'OIUfANN, ROTB, and ZUBOW leave no room for dOllbt Ihat the inerease of Ihe heat of combnslioll pe l' CH,-group of C. Hlo , C\H l .,

C7 H14 does not alllollllt 10 mOre than lIas been fOllnd in the pal'affills, wlticlt ollgltt to have beell th e case wilh increasillg lension .

This reslllt seems to be ill conlli ct witll tlle ex periences oblained about tho ring-closllres, Wlten Ih ere is no ring-tension pl'esent iJl the salul'I1led rings wilh more tll an six a loms, wlty are not Iltey fOl'med as easiJy as tlte five-rillgs, and wlty al'e th ey also cleal'ly less widely spread in natllre ?

This more dillicuh formation is, indeed, stl'ikiJlgly illu stl'ated by Ihe vel'y low constant of eq uilihrium of th e system propa ne diol 1.3 + acetone, especially whell compured with the eOl'l'esponding propane diol 1.2 + aeetone, the differellee of whicll rests on six­ring closlll'e against five-ring closure (see above),

An answer is easy to give also here, The pl'obability that the two OH-gronps in the propylene glycol

1.2 assume a favo\ll'able position witlt regard to the acetone molecule is so mnch gl'eater than for tlte 1,3 isomei'. In the pl'opane diol1 .2 a l''otat.ion of th e molecule parls round a single bond suffices, whel'eas in the pl'opane diol 1.3 the putlIs of tlle OH-gI'OllpS are mllch more involved and intl'icate, so that in Ilte same space of time they will certainly assume a fa vourable position much less frequently ,

We may sUlllmal'ise Ihe I'esu l't of Ihis st. lIdy as followR :

1. AI's0' at,(i)AlS th at are not drirredly b(!)Und t0 eaeh other ,. exereiBe

an aetion OT!' eaeh othet· in the same molecule. 2. 'Fhe satluated n@f1-cyclit' molet'll'les' execlde, among oU<tens,

mo-vements in w'hich the parts of the mt)~ecl1le l1evolve in opposite directiofl or' with diffel'ent veloci.ti:es l"ounC!l the silDigle bond·s· as·axes.

In case of not uniform load, as is tlre case in by flW the mast motec1t!1es (exceptecl' Rre only H" O2 , N, etc. C,H e, C,Cle, pet'haps o\t8llic acid, etc.) the movem:ents are it:regular, because tl.re m-est st~ble position of the atoms . will be passoo most fl"equently.

3. In the' satl1rat·ed ri'ng-shaped molecl1les witlr, six anei seven carbon atoms the I'ing-fornring atoms- are not fixed in one plane, but tIJer l~e tension-Iess in a curved sUl'face, vwhich tElwelsthrough space in und wlatory mo\'emen.ts.

Delft, October 1921.